Analysis of Cultural Participation by Two-Mode Three-Way Asymmetric Cluster Analysis and Multidimensional Scaling
摘要
Differences of cultural participation caused by the educational background and gender are investigated by analyzing relationships between nine cultural activities among seven groups having different social classes. The data consist of seven 9 × 9 similarity matrices (9 × 9 × 7 two-mode three-way similarities) where each matrix comes from a social group and is not necessarily symmetric. We extended ACLUSKEW (which can only deal with one-mode two-way asymmetric proximities) so that two-mode three-way asymmetric proximities can be dealt with and analyzed the data by the newly developed two-mode three-way asymmetric cluster analysis. We also analyzed the data by the two-mode three-way radius-distance model so that we can compare the results obtained by the cluster analysis and multidimensional scaling. The results show that asymmetric relationships between cultural activities are governed by the differences between highbrow activities and lowbrow activities and the amount of spending to participate in an activity whose objective is public. The results also show that the asymmetry of participating in cultural activities is more noticeable for male groups than for female groups; that asymmetry of relationships between cultural activities is smaller for groups with higher educational background than for those with lower educational background for the male groups.